archive.py 15.4 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
# Copyright 2018 Markus Scheidgen
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an"AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

15
16
17
18
19
"""
The archive API of the nomad@FAIRDI APIs. This API is about serving processed
(parsed and normalized) calculation data in nomad's *meta-info* format.
"""

20
from typing import Dict, Any
21
from io import BytesIO
22
23
import os.path
from flask import send_file
24
from flask_restplus import abort, Resource, fields
25
import json
26
import importlib
27

28
29
import nomad_meta_info

30
from nomad.files import UploadFiles, Restricted
31
from nomad import utils, search
32

33
from .auth import authenticate, create_authorization_predicate
Markus Scheidgen's avatar
Markus Scheidgen committed
34
from .api import api
35
from .repo import search_request_parser, add_query
36
from .common import calc_route, streamed_zipfile, pagination_model, build_snippet
37
38

ns = api.namespace(
39
40
    'archive',
    description='Access archive data and archive processing logs.')
41
42
43
44


@calc_route(ns, '/logs')
class ArchiveCalcLogResource(Resource):
45
    @api.doc('get_archive_logs')
46
    @api.response(404, 'The upload or calculation does not exist')
47
    @api.response(401, 'Not authorized to access the data.')
48
    @api.response(200, 'Archive data send', headers={'Content-Type': 'application/plain'})
49
    @authenticate(signature_token=True)
50
    def get(self, upload_id, calc_id):
51
52
53
        """
        Get calculation processing log.

54
        Calcs are references via *upload_id*, *calc_id* pairs.
55
        """
56
        archive_id = '%s/%s' % (upload_id, calc_id)
57

58
        upload_files = UploadFiles.get(
59
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
60

61
        if upload_files is None:
62
            abort(404, message='Upload %s does not exist.' % upload_id)
63
64
65

        try:
            return send_file(
66
                upload_files.archive_log_file(calc_id, 'rb'),
67
68
                mimetype='text/plain',
                as_attachment=True,
69
                cache_timeout=0,
70
71
                attachment_filename='%s.log' % archive_id)
        except Restricted:
72
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
73
74
        except KeyError:
            abort(404, message='Calculation %s does not exist.' % archive_id)
75
76
77
78


@calc_route(ns)
class ArchiveCalcResource(Resource):
79
    @api.doc('get_archive_calc')
80
    @api.response(404, 'The upload or calculation does not exist')
81
    @api.response(401, 'Not authorized to access the data.')
82
    @api.response(200, 'Archive data send')
83
    @authenticate(signature_token=True)
84
    def get(self, upload_id, calc_id):
85
86
87
        """
        Get calculation data in archive form.

88
        Calcs are references via *upload_id*, *calc_id* pairs.
89
        """
90
        archive_id = '%s/%s' % (upload_id, calc_id)
91

92
        upload_file = UploadFiles.get(
93
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
94

95
        if upload_file is None:
96
            abort(404, message='Archive %s does not exist.' % upload_id)
97
98
99

        try:
            return send_file(
100
                upload_file.archive_file(calc_id, 'rb'),
101
102
                mimetype='application/json',
                as_attachment=True,
103
                cache_timeout=0,
104
105
                attachment_filename='%s.json' % archive_id)
        except Restricted:
106
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
107
        except KeyError:
108
            abort(404, message='Calculation %s does not exist.' % archive_id)
109
110


111
112
113
114
115
archives_from_query_parser = search_request_parser.copy()
archives_from_query_parser.add_argument(
    name='compress', type=bool, help='Use compression on .zip files, default is not.',
    location='args')

116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
archives_from_query_model_fields = {
    'pagination': fields.Nested(pagination_model, skip_none=True),
    'scroll': fields.Nested(allow_null=True, skip_none=True, model=api.model('Scroll', {
        'total': fields.Integer(description='The total amount of hits for the search.'),
        'scroll_id': fields.String(allow_null=True, description='The scroll_id that can be used to retrieve the next page.'),
        'size': fields.Integer(help='The size of the returned scroll page.')})),
    'results': fields.List(fields.Raw, description=(
        'A list of search results. Each result is a dict with quantities names as key and '
        'values as values')),
    'archive_data': fields.Raw(description=('A dict of archive data with calc_ids as keys ')),
    'code_snippet': fields.String(description=(
        'A string of python code snippet which can be executed to reproduce the api result.')),
}
for group_name, (group_quantity, _) in search.groups.items():
    archives_from_query_model_fields[group_name] = fields.Nested(api.model('ArchiveDatasets', {
        'after': fields.String(description='The after value that can be used to retrieve the next %s.' % group_name),
        'values': fields.Raw(description='A dict with %s as key. The values are dicts with "total" and "examples" keys.' % group_quantity)
    }), skip_none=True)
archives_from_query_model = api.model('RepoCalculations', archives_from_query_model_fields)


@ns.route('/download')
class ArchiveDownloadResource(Resource):
139
140
    manifest_quantities = ['upload_id', 'calc_id', 'external_id', 'raw_id', 'pid', 'calc_hash']

141
    @api.doc('archive_zip_download')
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
    @api.expect(archives_from_query_parser, validate=True)
    @api.response(200, 'File(s) send', headers={'Content-Type': 'application/zip'})
    @authenticate(signature_token=True)
    def get(self):
        """
        Get calculation data in archive form from all query results.

        See ``/repo`` endpoint for documentation on the search
        parameters.

        Zip files are streamed; instead of 401 errors, the zip file will just not contain
        any files that the user is not authorized to access.

        The zip file will contain a ``manifest.json`` with the repository meta data.
        """
        try:
            args = archives_from_query_parser.parse_args()
            compress = args.get('compress', False)
        except Exception:
            abort(400, message='bad parameter types')

        search_request = search.SearchRequest()
        add_query(search_request, search_request_parser.parse_args())

167
        calcs = search_request.execute_scan(order_by='upload_id')
168
169

        def generator():
170
            manifest = {}
Markus Scheidgen's avatar
Markus Scheidgen committed
171
            upload_files = None
172
173
174
            for entry in calcs:
                upload_id = entry['upload_id']
                calc_id = entry['calc_id']
Markus Scheidgen's avatar
Markus Scheidgen committed
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
                if upload_files is None or upload_files.upload_id != upload_id:
                    if upload_files is not None:
                        upload_files.close_zipfile_cache()

                    upload_files = UploadFiles.get(
                        upload_id, create_authorization_predicate(upload_id))

                    if upload_files is None:
                        utils.get_logger(__name__).error('upload files do not exist', upload_id=upload_id)
                        continue

                    upload_files.open_zipfile_cache()

                yield (
                    '%s.%s' % (calc_id, upload_files._archive_ext), calc_id,
                    lambda calc_id: upload_files.archive_file(calc_id, 'rb'),
                    lambda calc_id: upload_files.archive_file_size(calc_id))
192

193
                manifest[calc_id] = {
194
                    key: entry[key]
195
                    for key in ArchiveDownloadResource.manifest_quantities
196
197
                    if entry.get(key) is not None
                }
198

Markus Scheidgen's avatar
Markus Scheidgen committed
199
200
201
            if upload_files is not None:
                upload_files.close_zipfile_cache()

202
203
204
205
206
207
208
209
210
211
212
213
            try:
                manifest_contents = json.dumps(manifest).encode('utf-8')
            except Exception as e:
                manifest_contents = json.dumps(
                    dict(error='Could not create the manifest: %s' % (e))).encode('utf-8')
                utils.get_logger(__name__).error(
                    'could not create raw query manifest', exc_info=e)

            yield (
                'manifest.json', 'manifest',
                lambda *args: BytesIO(manifest_contents),
                lambda *args: len(manifest_contents))
214

215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
        return streamed_zipfile(
            generator(), zipfile_name='nomad_archive.zip', compress=compress)


@ns.route('/query')
class ArchiveQueryResource(Resource):
    @api.doc('archive_json_query')
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
    @api.expect(search_request_parser, validate=True)
    @api.marshal_with(archives_from_query_model, skip_none=True, code=200, description='Search results sent')
    @authenticate(signature_token=True)
    def get(self):
        """
        Get archive data in json format from all query results.

        See ``/repo`` endpoint for documentation on the search
        parameters.

        The actual data are in archive_data and a supplementary python code to execute
        search is wirtten in code_snippet.
        """
        try:
            args = search_request_parser.parse_args()
            scroll = args.get('scroll', False)
            scroll_id = args.get('scroll_id', None)
            page = args.get('page', 1)
            per_page = args.get('per_page', 10 if not scroll else 1000)
            order = args.get('order', -1)
            order_by = 'upload_id'
        except Exception:
            abort(400, message='bad parameter types')

        try:
            assert page >= 1
            assert per_page > 0
        except AssertionError:
            abort(400, message='invalid pagination')

        if order not in [-1, 1]:
            abort(400, message='invalid pagination')

        search_request = search.SearchRequest()
        add_query(search_request, search_request_parser.parse_args())

        try:
            if scroll:
                results = search_request.execute_scrolled(scroll_id=scroll_id, size=per_page)

            else:
                results = search_request.execute_paginated(
                    per_page=per_page, page=page, order=order, order_by=order_by)

        except search.ScrollIdNotFound:
            abort(400, 'The given scroll_id does not exist.')
        except KeyError as e:
            import traceback
            traceback.print_exc()
            abort(400, str(e))

        # build python code snippet
        snippet = build_snippet(args, os.path.join(api.base_url, ns.name, 'query'))
        results['code_snippet'] = snippet

        data = {}
        calcs = results['results']
        try:
            upload_files = None
            for entry in calcs:
                upload_id = entry['upload_id']
                calc_id = entry['calc_id']
                if upload_files is None or upload_files.upload_id != upload_id:
                    if upload_files is not None:
                        upload_files.close_zipfile_cache()

                    upload_files = UploadFiles.get(
                        upload_id, create_authorization_predicate(upload_id))

                    if upload_files is None:
                        raise KeyError

                    upload_files.open_zipfile_cache()

                fo = upload_files.archive_file(calc_id, 'rb')
                data[calc_id] = json.loads(fo.read())

            if upload_files is not None:
                upload_files.close_zipfile_cache()

        except Restricted:
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))

        except KeyError:
            abort(404, message='Calculation %s/%s does not exist.' % (upload_id, calc_id))

        results['archive_data'] = data

        return results, 200
312
313


314
315
@ns.route('/metainfo/<string:metainfo_package_name>')
@api.doc(params=dict(metainfo_package_name='The name of the metainfo package.'))
316
317
318
319
class MetainfoResource(Resource):
    @api.doc('get_metainfo')
    @api.response(404, 'The metainfo does not exist')
    @api.response(200, 'Metainfo data send')
320
    def get(self, metainfo_package_name):
321
322
323
324
        """
        Get a metainfo definition file.
        """
        try:
325
            return load_metainfo(metainfo_package_name), 200
326
        except FileNotFoundError:
327
            parser_prefix = metainfo_package_name[:-len('.nomadmetainfo.json')]
328

329
            try:
330
331
332
                return load_metainfo(dict(
                    parser='%sparser' % parser_prefix,
                    path='%s.nomadmetainfo.json' % parser_prefix)), 200
333
334
            except FileNotFoundError:
                abort(404, message='The metainfo %s does not exist.' % metainfo_package_name)
335
336
337
338
339


metainfo_main_path = os.path.dirname(os.path.abspath(nomad_meta_info.__file__))


340
341
342
def load_metainfo(
        package_name_or_dependency: str, dependency_source: str = None,
        loaded_packages: Dict[str, Any] = None) -> Dict[str, Any]:
343
344
345
346
347
    """
    Loads the given metainfo package and all its dependencies. Returns a dict with
    all loaded package_names and respective packages.

    Arguments:
348
349
        package_name_or_dependency: The name of the package, or a nomadmetainfo dependency object.
        dependency_source: The path of the metainfo that uses this function to load a relative dependency.
350
351
352
353
354
355
        loaded_packages: Give a dict and the function will added freshly loaded packages
            to it and return it.
    """
    if loaded_packages is None:
        loaded_packages = {}

356
357
    if isinstance(package_name_or_dependency, str):
        package_name = package_name_or_dependency
358
        metainfo_path = os.path.join(metainfo_main_path, package_name)
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
    else:
        dependency = package_name_or_dependency
        if 'relativePath' in dependency:
            if dependency_source is None:
                raise Exception(
                    'Can only load relative dependency from within another metainfo package')

            metainfo_path = os.path.join(
                os.path.dirname(dependency_source), dependency['relativePath'])

        elif 'metainfoPath' in dependency:
            metainfo_path = os.path.join(metainfo_main_path, dependency['metainfoPath'])

        elif 'parser' in dependency:
            parser = dependency['parser']
            path = dependency['path']
            try:
                parser_module = importlib.import_module(parser).__file__
            except Exception:
                raise Exception('Parser not installed %s for metainfo path %s' % (parser, metainfo_path))

            parser_directory = os.path.dirname(parser_module)
            metainfo_path = os.path.join(parser_directory, path)

        else:
            raise Exception('Invalid dependency type in metainfo package %s' % metainfo_path)

        package_name = os.path.basename(metainfo_path)
387
388
389
390
391
392
393
394
395
396
397
398

    package_name = os.path.basename(package_name)

    if package_name in loaded_packages:
        return loaded_packages

    with open(metainfo_path, 'rt') as f:
        metainfo_json = json.load(f)

    loaded_packages[package_name] = metainfo_json

    for dependency in metainfo_json.get('dependencies', []):
399
        load_metainfo(dependency, dependency_source=metainfo_path, loaded_packages=loaded_packages)
400
401

    return loaded_packages